About Me

Hello! I’m Justin Donaldson, a passionate Human-Computer-Interaction/Data/AI Scientist and Engineer with a diverse skill set spanning large language model research, data science, and machine learning. My journey has led me from academia to industry, and back again. I’m always eager to learn something complex, and turn it into something beautiful.

Currently

CTO : Hushh

I’m helping start Hushh.ai, a company I co-founded with Manish Sainani in 2024. At Hushh, we’re pioneering on-device hybrid LLM search capabilities and nurturing a vibrant R&D culture with a talented team of international interns.

Investor/Advisor

Dagworks

I met Stefan and Elijah when they were just getting started building Dagworks.io. They had some of the same notions I had on the importance of Directed Acyclic Graphs as a fundamental programming theme. Their decorator-based approach for static methods creates the necessary coupling between imperative graph steps, and overall graph execution order in an especially elegant way. I believe most people will adopt it at some point if they are using Python to manage data. It’s just that good.

PulseSpace

If you want to give your company the best chance for success, you have to think outside the box in order to network. I met Karl Stedman when he was taking me to the airport as a Lyft driver. There’s an awful lot you can learn about someone in a 30 minute drive to the airport, and I got a chance to hear about Litepulse (soon to be PulseSpace), and their goal of providing remote power for satellites. With the growth of the private space industry thanks to SpaceX and Blue Origin, there’s a whole ecosystem of young startups looking to provide service level agreements for orbital infrastructure. PulseSpace is looking to be a sort of “gas station” for satellites that may have malfunctioning or damaged solar panels, or maybe they just need some extra juice. Even though I don’t have a background in astrophysics, I think they’re bound to run into all sorts of gnarly software problems. I’m happy to help however I can there.

Lately

Prior to Hushh.ai, I spent a rewarding decade at Salesforce, where I

wore multiple hats as a Principal Data Scientist and Engineer. I had the privilege to lead early search infrastructure modernization projects like migrating the core search infrastructure to model-trained ranking coefficients , and also developed patented deep learning models for detecting error states in Salesforce’s enormous code base. I co-founded the Service Cloud Datascience team and contributed significantly to the creation of an ML education program for engineers. My contributions also extended to investigating and developing GPT-3 technologies.

Earlier in my career, I co-founded BigML and served as its President from 2011 to 2013. There, I spearheaded the development of interactive visualizations and took charge of various administrative tasks, setting the foundation for the company’s growth.

What I know

My expertise encompasses model optimization, time series analysis, anomaly detection, and advanced ML/web visualization. I’m well-versed in programming languages like Python, R, and familiar with Docker and core web technologies. Over the years, I’ve had the opportunity to work on diverse and impactful projects with trusted Salesforce partners, crafting custom solutions ranging from advanced candidate profilers to anomaly detection models for global enterprises. Make sure to check out my C.V. for more details.

Let’s Connect!

I’m always eager to connect with like-minded professionals and enthusiasts. Feel free to reach out to me via: email, Twitter, LinkedIn, or explore my CV and GitHub

Looking forward to connecting with you!